1. Field of the Invention
The invention relates to a multi-user computer system, in particular to a method of controlling the creation of a user session in such a system.
2. Brief Description of the Prior Art
Multi-user computer environments are known in a number of different variants. Common examples of such environments consist of one or more servers, and a number of computer terminals connected to the server, usually through a network. A server operating system is installed on the server(s), together with several applications, such as spreadsheet programs, word processors, etc. The software on the server, and other resources in the system are accessible to users of the terminals. To access the software and other resources, the user makes a request to log on to the system. Such a request marks the start of the creation of a user session. This entails amongst others the loading of device drivers, setting up background processes specified for the user, and creating page files, for example. The creation of a user session is a process that makes heavy use of the server resources, particularly the disk drives on which the various executables are stored.
System management programs exist to manage user sessions and their creation, in order, amongst others, to ensure that the system is not overloaded. This frequently occurs at certain times, for example at the start of the working day, when all employees in an organisation log on at once. At such times, the system can become very slow, with the creation of user sessions taking an inordinate amount of time. In the worst case, the system can break down all together. One type of system management program uses the principle of load balancing to alleviate problems associated with many simultaneous log-on requests. The processing of these requests is spread out as evenly as possible amongst a group of servers, each capable of independently creating user sessions. This principle suffers from the disadvantage that an increase in the number of users inevitably requires the addition of extra servers to the system at some point. The system is thus dimensioned to cope with peak use, since subsequent use of the system is usually much less heavy, due to users pausing with their work, for example.